The current proposal aims to facilitate hand function in chronic stroke patients. Deficits in hand function are extremely common after stroke and persist for a prolonged period of time. While conventional techniques have emphasized the repetitive practice of simple movements, the role of coordination in rehabilitation has not been explored extensively. In this proposal, the PIs examine the influence of training novel finger coordination patterns as a means of improving hand function and dexterity. They propose an innovative protocol to induce the reorganization of finger movements. Instead of providing explicit instructions, the PIs will elicit different coordination patterns in the context of a simpe task. Participants wear a data glove and the signals from the glove are mapped on to the position of a computer cursor. The goal of the participants is to move their fingers so that they learn to move the computer cursor to different targets on a screen. This protocol has the advantage of reducing the complex task of finger coordination into the simpler task of planning cursor movements. The first aim investigates how changing the relation between finger and cursor motion (i.e., changing the map) affects the reorganization of finger coordination. Specifically, the PIs examine the effectiveness of changing two features of the map (gain and variability) in driving the participants to reorganize their finger movements. The second aim examines the feasibility of using this protocol to improve hand function and dexterity in chronic stroke patients. Patients undergo training in the glove task with different maps designed to elicit varied finger movements. Clinical tests are then used to determine if training resulted in improved hand function. The results of these experiments have the potential to transform rehabilitation approaches by introducing the exploration of novel coordination patterns as an important component of training protocols.